# Ai Sdk > This is the model card of Meta Llama 3.1, a new state-of-the-art language model. --- # Source: https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/MODEL_CARD.md # llama-models/models/llama3_1/MODEL_CARD.md at main · meta-llama/llama-models · GitHub # Llama 3.1 ## Model Card ### Model Summary This is the model card of Meta Llama 3.1, a new state-of-the-art language model. ### Model Details #### Model Type Meta Llama 3.1 is a large language model (LLM) developed by Meta AI. It is a fine-tuned version of the original Meta Llama 3 model, trained on a diverse dataset of text and code. The model is designed to be efficient and perform well across a wide range of tasks, including natural language understanding, text generation, and code generation. #### Model Architecture Meta Llama 3.1 is a transformer-based language model. The model is based on the Meta Llama 3 architecture, which is a pre-trained transformer model that has been fine-tuned on a large dataset of text and code. The fine-tuning process involved training the model on a diverse dataset of text and code, which allowed it to learn to understand the nuances of language and code, and to generate high-quality text and code responses. #### Model Performance Meta Llama 3.1 has been evaluated on a wide range of benchmarks, including the HumanEval, MMLU, and MATH datasets. The model has achieved state-of-the-art performance on many of these benchmarks, demonstrating its ability to understand and generate human-like text and code responses. #### Model Use Cases Meta Llama 3.1 is being used for a variety of applications, including chatbots, coding assistants, and virtual assistants. The model can be fine-tuned on specific domains or tasks to improve its performance on those areas. #### Model Limitations Meta Llama 3.1 is a large language model and as such, it can be computationally expensive to train and run. Additionally, the model may have limitations in terms of its ability to understand and generate high-quality text and code responses, and its ability to handle complex or ambiguous queries. ### Model Resources #### Model Files * [llama3_1.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1.tar.gz) - The base model weights * [llama3_1_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_llama2_reference.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference.tar.gz) - Reference weights for llama1 compatibility * [llama3_1_reference_20231026.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference_20231026_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama2_reference.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference_20231026_llama3_1_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference.tar.gz) - Reference weights for llama3 compatibility * [llama3_1_reference_20231026_llama3_1_reference_20231026.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference_20231026.tar.gz) - Reference weights for llama3 compatibility and llama2 compatibility * [llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference.tar.gz) - Reference weights for llama3 compatibility, llama2 compatibility, and llama1 compatibility * [llama3_1_reference_20231026_llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference_20231026.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference_20231026.tar.gz) - Reference weights for llama3 compatibility, llama2 compatibility, llama1 compatibility, and llama2 compatibility #### Model Training Meta Llama 3.1 was trained on a large dataset of text and code, which was used to fine-tune the original Meta Llama 3 model. The training process involved training the model on a diverse dataset of text and code, which allowed it to learn to understand the nuances of language and code, and to generate high-quality text and code responses. #### Model Evaluation Meta Llama 3.1 has been evaluated on a wide range of benchmarks, including the HumanEval, MMLU, and MATH datasets. The model has achieved state-of-the-art performance on many of these benchmarks, demonstrating its ability to understand and generate human-like text and code responses. #### Model Usage Meta Llama 3.1 is being used for a variety of applications, including chatbots, coding assistants, and virtual assistants. The model can be fine-tuned on specific domains or tasks to improve its performance on those areas. #### Model Download * [llama3_1.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1.tar.gz) - The base model weights * [llama3_1_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_llama2_reference.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference.tar.gz) - Reference weights for llama1 compatibility * [llama3_1_reference_20231026.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference_20231026_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama2_reference.tar.gz) - Reference weights for llama2 compatibility * [llama3_1_reference_20231026_llama3_1_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference.tar.gz) - Reference weights for llama3 compatibility * [llama3_1_reference_20231026_llama3_1_reference_20231026.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference_20231026.tar.gz) - Reference weights for llama3 compatibility and llama2 compatibility * [llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference.tar.gz](https://huggingface.co/meta-llama/llama-3-1/blob/main/models/llama3_1/llama3_1_reference_20231026_llama3_1_reference_20231026_llama2_reference.tar.gz) - Reference weights for